The great demand from basic science and drug design for detailed, atomic level information about the structure of proteins and nucleic acids resulted in a large-scale experimental effort, with crystallography being its predominant technique. The aim of this grant is to improve methods of crystallographic diffraction data analysis in the areas of data collection, integration of diffraction peaks, scaling and merging. The methods address both the mainstream use and the frontiers of crystallography. Improvements to the core of current methods will result from the precise modeling of diffraction peaks' shape, the removal of diffraction artifacts, the corrections for radiation damage-induced effects and the better diagnostics of possible problems, like pseudo symmetry and/or twinning etc. The challenging structural projects require complex optimization of crystallization and cryo-protection conditions. Such projects benefit greatly from averaging, if possible, the data from multiple crystals, so far mostly avoided due to lack of appropriate tools. We plan to develop a novel type of diffraction signal analysis that will compare data from multiple crystals to define their diffraction quality and consistency. Data from multiple crystals will be decomposed into statistically significant signals, including the desired phasing signals and the description of non-isomorphism. The results of this analysis will be straightforward to interpret and optimal in terms of averaging data from not necessarily exactly isomorphous crystals. This development is particularly significant due to the technology scaling up the production of crystals and the collection of diffraction data. The improved effectiveness of the experimental work will result also from the expanded experiment planning tools, automatic interactions with external databases, data management tools and parallel computing. The developed algorithms and methods will be implemented in the HKL2000 suite, widely used by the crystallographic community. [unreadable] [unreadable] [unreadable]